Without a doubt, the latest crop of devices from Apple are set to perform the best on the latest OS update even if it’s no longer the original OS the Mac shipped with. While it can be challenging to imagine or quantify precisely how fast macOS Tahoe will be upon it’s debut to the public, some artificial benchmarks can give us a sort-of an idea.
CPU Benchmarks
When comparing the 15″ M4 MacBook Air from when it was announced to the debut of the developer beta, GeekBench 6 gives the appearance of increased performance.
- See: Benchmark results on Geekbench website
. - On macOS Tahoe: 3668 (Single Core) 15225 (Multicore)
- On macOS Sequoia: 3638 (Single Core) 14310 (Multicore)
In nearly every category, single core performance was either about the same or better, whereas every multicore category saw improved performance across the board.
.
GPU Performance
GPU Performance isn’t as telling, although it is remarkable to see how GPU performance doesn’t take a major hit (if at all) despite the major changes in the UI. This is a bright spot on the newer OS considering how these changes won’t hamper cutting edge tech just because “it can do something that older tech can’t” – but instead leverage it.
Of course, an OCLP patched Mac running Tahoe will be able to run these elements especially if it has AVX2 with a supported GPU. But seeing as those are older machines which require stronger GPUs to leverage the same UI elements, these changes will be more affect-laden for those.
OpenCL
- On macOS Tahoe: 35372
- On macOS Sequoia: 35494
- (See Comparison on website)
Metal
- On macOS Tahoe: 54314
- On macOS Sequoia: 54999
- (See Comparison on website)
It makes sense how in the Metal GPU artificial benchmarks, particle physics and background blur took a slight hit in performance when not in game mode or running anything else – this is the OS using more GPU power in reference to Liquid Glass. On the other hand, there is increased performance in face detection.
.
In Conclusion
The numbers tell a story of a minor potential change and these benchmarks can swing just as far in the negative direction given the right circumstances. Perhaps, however, there is a grain of truth in all this, as some early adopters online are reporting a far better experience.
“Everything feels more faster from searching in Settings to bringing up an emoji picker.” as sai @heysaik says it in a post on X.